DO NOT TOP POST Am 31.05.2012 11:32, schrieb Roland Roland: > On 5/31/12 11:23 AM, Reindl Harald wrote: >> >> Am 31.05.2012 09:13, schrieb Roland RoLaNd: >>> Dear all,I have a task to anonymize data in 82000 >>> record (for the time being)i've tested my script against >>> about 30 of them, and it takes about 2.4 sec for each >>> query to be executed. >> provide table structure and query example >> sounds like a bad design without key >> >> how can 30 updates take 2.4 seconds? >> >>> i'm seeking help with the following: >>> - what's the best way to run such a script without affecting the DB server >>> performance? >>> Should i limit the script to implement N number of records at a time? >>> and then sleep or is there a better way? >>> - What's the best practice of handling errors and warnings in such a >>> situation? >>> - How can i prevent the runtime errors and mysql locks? >> wrong question >> >> [--] Data in InnoDB tables: 6G (Tables: 49) >> [--] Up for: 5d 11h 17m 15s (493M q [1K qps], 43K conn, TX: 38B, RX: 14B) >> [--] Reads / Writes: 89% / 11% >> >> as you can see 30 queries or even 88.000 queries must not >> make any trouble if you table-design is OK > > i agree it wouldn't cause trouble, though it might lock mysql as > there's a number of other databases running on the same server. > so performance is an issue even if it's just a CPU/RAM peak.
if your queries are fast enough that should not matter as long 30 updates take more than 2 seconds something goes terrible wrong on your machine as said: if you need help post informations! * mysql-configuration * hardware-specs (especially RAM) * table definition * query example you can also use "EXPLAIN" to test your queries and see what is going wrong: http://dev.mysql.com/doc/refman/5.0/en/explain.html as long you do not provide specific informations nobody will and can help you - there is no magical "do thsi to get best performance" anywhere in the IT
signature.asc
Description: OpenPGP digital signature
